AMD Radeon RX 7900 GRE 16GB vs AMD Radeon PRO W7900 48GB

Both GPUs launched in 2023: the Radeon RX 7900 GRE on RDNA 3 and the Radeon PRO W7900 on RDNA 3.

Dense throughput for the Radeon RX 7900 GRE against the Radeon PRO W7900: FP32 46 vs 61 TFLOPS, FP16 92 vs 123 TFLOPS. Sparse figures, where the vendor publishes them, appear under each dense number in the performance table.

Memory is 16 GB against 48 GB, fed at 576 GB/s versus 864 GB/s. More memory per GPU means larger models fit before you have to shard across cards, which often matters more than raw TFLOPS for inference.

Power budgets are 260 W for the Radeon RX 7900 GRE and 295 W for the Radeon PRO W7900. At FP32 that works out to 0.18 against 0.21 TFLOPS per watt.

For LLM training and inference, weight the FP8 and FP16 rows and memory capacity most heavily. For scientific and HPC workloads, the FP64 row is the one to read.